274 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 51 



altitude at which the water particles reverse their motion and begin 

 to fall in consequence of the increased size to which they have 

 attained by reason of the progressive condensation attending their 

 long path. 



Rather will this current continue far above the possible boundary 

 of the cloud not only in the case of the summer cumuli proper, but in 

 all cases where we have to do with the ascension of individual masses 

 of air and their penetration into upper strata. 



But in fact in the case of the heat thunderstorms with their 

 perpetually renewed whirl about a horizontal axis we have to do 

 in a certain sense with a steady process even if, as such, it is moving 

 forward . 



In this current of air flowing outward from a "thunderhead" 

 condensation must again occur in consequence of the progressive 

 cooling. 27 But in this case, on the one hand, the quantity of aque- 

 ous vapor coming into consideration is slight; on the other hand, 

 the precipitation will be directly in the form of crystals of ice or 

 snow because of the temperature prevailing at these altitudes. 



In this process which is comparable with sublimation, the super- 

 saturation or the subcooling can no longer play any important role 

 even when that might otherwise be possible because of the small 

 quantity of water remaining present. Hence in these clouds there 

 do not exist the protruding and expanding heads, but in accord 

 with the steadily ascending current they develop rather in the screen- 

 like form of a layer of cirrus. 



Doubts have indeed been expressed as to whether the screen of 

 cirrus that accompanies the thunderstorm cloud really consists 

 always of ice or snow, since the characteristic optical phenomena 

 are not always observable in it. Hence it seems to me to be impor- 

 tant to point out that between ice clouds at comparatively low 

 altitudes, such as correspond to the layer of cirrus, and those in the 

 highest layers of the atmosphere, there may be very considerable 

 differences that may exert an influence on the optical behavior. 28 

 Thus, at very low temperatures ice needles easily form, whereas at 

 temperatures that lie near the freezing point the precipitation takes 

 place in the form of starlike crystals of snow or indeed masses of 

 snow, which latter must be less conducive to the development of the 

 well-known optical phenomena. 



27 Compare pp. 237 et seq. of the preceding collection of translations. 



28 According to recent investigations it must be recognized as very prob- 

 able that the ions play an important part in the formation of the cirrus screen 

 (overflow or false cirrus or cirrus veil or cirrus screen) as indeed in the for- 

 mation of cirri in general. (Note added in 1905. W. v. B.) 



